ENVIRONMENTAL APPLICATIONS OF CHIRAL HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY 1W-Y. Lee and 2J.M. Salvador 1Environmental Science and Engineering, 2Department of Chemistry, University of Texas at El Paso, El Paso, TX 79968; Phone: (915) 747-5704; FAX: (915) 747-5748. ABSTRACT A commercial pharmaceutical analysis chiral method development kit (Chirex Column Kit A, Phenomenex) was used to analyze six pesticide stereoisomer mixtures. The pesticides were selected from the 1994-1995 National Water Quality Assessment Program (NAWQA) covering New Mexico and Texas. Three stereoisomer mixtures were separable on three different columns. The instrumental detection limit and detection linear range of chiral high performance liquid chromatography (HPLC) of these analytes were determined. The efficient extraction (> 90%) from soil of one chiral mixture was demonstrated. The resolution of a shorter kit column versus an analytical column was also compared. Key words: chiral, chromatography, HPLC, pesticide, stereoisomer

INTRODUCTION limited to only one enantiomer, with the activity More than 1.1 billion pounds of pesticides of the other enantiomer being less effective, are used in the United States each year to inactive, or different (Ariens et al., 1988; Buser control weeds, insects, and pests in agricultural and Müller, 1995; Venis, 1982; Faller et al., and non-agricultural settings (Barbash and 1991; Renner, 1996). For example, some Resek, 1996). The increasing use of synthetic bacteria only degrade one enantiomer of organic pesticides raises many concerns about Mecoprop (R) (Tedd et al., 1994; Ludwig et potential adverse effects on the environment and al., 1992; Falconer et al., 1995; Iwata et al., human health, and increases the importance of 1998). The toxicity of the (R)-enantiomer of analyzing pesticides in the soil, water, and air Fonofos, an organophosphorus pesticide, in (Barceló, 1991; Müller and Buser, 1995). To mice is greater than the toxicity of its mirror assess the environmental impact of pesticides image, (S)-Fonofos (Kurihara et al., 1997). and reduce the risk of public exposure, all Ignoring the existence of enantiomers can lead aspects of pesticide chemistry need to be well to incorrect toxicological, distribution, and understood and studied. degradation data. Ongoing research is aimed at Approximately a quarter of the hundreds identifying chiral pesticides, determining their of pesticides in use are asymmetric or chiral distribution, and measuring their degradation. (Milne, 1995). Chiral molecules may exist as Chromatography is the most common mixtures of non-superposable mirror images or method for enantiomeric analysis. Gas chroma- enantiomers. Enantiomers have different tography (GC) is the primary method used in chemical and physical properties in asymmetric pesticide analyses (Grosser et al., 1993). environments. The desired biological activity of However, high-performance liquid chromatog- a chiral pesticide enantiomer mixture may be raphy (HPLC) is more suitable for larger, non-

36 Proceedings of the 2000 Conference on Hazardous Waste Research cides are chiral. The stereoisomer mixtures of Fonofos, Malathion, Metolachlor, Napropamide, Permethrin, and Propargite are shown in Figure 1. Although the enantiomers of chiral pesti- cides may have different biological properties in the environment, the NAWQA study completely ignored stereochemistry. This may be due to either a lack of knowledge of the potential problem or a lack in methods or technology to analyze chiral pesticides. Current U.S. EPA methods (507 and 508) use GC to detect some of these pesticides in drinking water but do not separate stereoisomers of chiral pesticides. For Figure 1. Structures of the stereoisomers of our work we selected six chiral pesticides from the six chiral pesticides from the NAWQA the NAWQA study to demonstrate the applica- study of Texas and New Mexico. tion of chiral HPLC in pesticide analysis. volatile, polar, and thermally labile pesticides Though one of the seven chiral pesticides (Buser and Müller, 1995; Farran et al., 1996). studied in NAWQA, α-Lindane was excluded Also, HPLC can tolerate large-volume injec- because our HPLC detector depends on UV tions of aqueous samples, rendering it ideal for (ultra violet-visible) absorption. screening water samples (Grosser et al., 1993). Since no single chromatography column In fact, the number of official U.S. Environmen- can separate all compounds, to find suitable tal Protection Agency (EPA) methods using stationary phases and separation conditions can HPLC grew to more than 40 approved and be expensive and time consuming. Thus we draft methods by 1993 (Grosser et al., 1993). used the shorter columns of a commercial chiral The Rio Grande Valley study unit of the HPLC method development kit to economically U.S. Geological Survey National Water Quality survey the efficiency of normal and reverse- Assessment Program (NAWQA) conducted a phase systems. The survey was used to guide study of the occurrence and distribution of the purchase of more expensive and longer pesticides in the surface water of our region of analytical columns. New Mexico and Texas (Healy, 1996). The We used Soxhlet extraction (EPA Solid study selected 40 pesticides and their metabo- Waste Test Methods, SW-846 Method 3540C lites on the basis of national usage, and the (Understanding Environmental Methods, CD- development and cost-effectiveness of analytical ROM, 1998) to demonstrate the recovery and procedures. Seven of the 40 selected pesti- analysis of a chiral pesticide mixture from soil

Proceedings of the 2000 Conference on Hazardous Waste Research 37 samples. This method is applicable to the ses were performed using Winner on Windows quantitative extraction of nonvolatile and (WOW) software. semivolatile organic compounds from solids The Chirex™ Chiral Method Develop- such as soil, relatively dry sludge, and solid ment Kit A (Phenomenex, Inc., Torrance, waste in preparation for a variety of chromato- Calif.) used in this work contained five 50 x 3.2 graphic procedures. Soxhlet extraction uses mm normal and reverse-phase HPLC columns: relatively inexpensive glassware, once loaded Chirex™ 3001, 3005, 3010, 3014, and 3020. requires little hands-on manipulation, and An additional Chirex™ 3001 analytical size provides efficient extraction. However, it uses (250 x 4.6 mm) column was also purchased. fairly large volumes of solvent and is rather time All HPLC injections consisted of 1µL into a consuming (16 to 24 hours). Combined with 20µL loop. SW-846 Method 3540C, the chiral HPLC Extraction and Concentration analysis methods presented are a beginning The Soxhlet extractor (Kimax, Fisher to study the different bioactivity, biodegra- Science) had a 40 mm ID with a 500-mL round dation, accumulation, and distribution of bottom flask. Soil samples were contained in pesticide enantiomer pairs in soil, surface cellulose thimbles (Whatman, Fisher Science). water, and groundwater. A Kuderna-Danish (K-D) apparatus was EXPERIMENTAL used to concentrate extracts. It included a 10- mL graduated concentrator tube, a 500-mL Materials evaporation flask attached to the concentrator Racemates (1:1 enantiomer mixtures) of tube with springs, and a Snyder column. A Metolachlor, Napropamide, and Permethrin solvent vapor recovery system (Kontes, Fisher were kindly provided by Norvatis (Greens- Science) was attached to the top of the Snyder boro, N.C.) and Zeneca (Richmond, Calif.). column to reduce emissions and minimize waste. Mixtures of Fonofos, Propargite, and Malathion were purchased from Chem Service Methods (West Chester, PA.). HPLC grade solvents Preparation of standards (isopropanol, hexane, acetonitrile, and water) To prepare stock standard solutions, 25 were obtained from Aldrich Chemical Co., mg of each pesticide were accurately weighed Milwaukee, Wis. All materials were used and dissolved in an appropriate solvent to make without further purification. a 25-mL solution. The purity of each pesticide Chromatography was 96% or greater; therefore, weights were Liquid chromatography was performed used without correction to calculate the concen- using a Spectra-Physics HPLC system consist- tration of each stock standard solution. Methyl- ing of a P2000 gradient pump and a UV2000 ene chloride was used to dissolve detector. Data acquisition, storage, and analy- Napropamide, and hexane was used to dissolve

38 Proceedings of the 2000 Conference on Hazardous Waste Research ROM, 1998). Soil was collected, passed through a 100-mesh screen, washed with water and methanol, and air dried in a hood overnight. A 10-g soil sample was then accurately weighed and put in an oven overnight at 105 °C. The percent dry weight was determined by the following equation:

g of dry sample % dry weight=X 100% g of sample

The oven-dried soil sample and 10 g of anhydrous sodium sulfate were mixed and put into a thimble. Soxhlet extraction of the soil sample was then performed as outlined in Figure 2. An outlined procedure of Soxhlet Figure 2. extraction of soil samples. For reasons to be discussed later, Napropamide was chosen for soil sample Fonofos, Malathion, Metolachlor, Permethrin, recovery analysis. A 50-mL spiking standard and Propargite. In the cases that separation of solution was prepared by dissolving 50 mg of isomers was not observed by HPLC, no pure Napropamide in methanol. The stock further standard solutions were made. Other- standard solution was stored in a Teflon-sealed wise 10 mL of standard solutions of each container at 4 °C. pesticide in 0.1, 0.5, 1.0, 5, 10, 50, 100, 250, Since there is no recommendation for and 500 ppm concentrations were prepared Napropamide analysis, naphthalene was chosen from their stock solution for linear-detection as a surrogate, a compound that is chemically range determination. similar to the analyte but is not expected to Instrumental detection limits (IDL) occur in an environmental sample. About 50 The HPLC UV detector was set to a 210- mg of pure naphthalene were added to a 50-mL nm wavelength for strong absorption by all six volumetric flask and diluted with methanol to analyzed mixtures. For each standard solution, make a 50-mL surrogate standard solution. triplicate HPLC analyses were performed at Adding 1.0 mL of the surrogate standard and room temperature. 1.0 mL of spiking solution into the soil sample Soil Recovery analysis gave a final concentration of both surrogate and Local soil samples were conditioned spiking standards of 100-mg/Kg soil. according to EPA SW-846 Method 3500 This soil sample was extracted in a Soxhlet (Understanding Environmental Methods, CD- apparatus with 250 mL of methylene chloride

Proceedings of the 2000 Conference on Hazardous Waste Research 39 for 24 hours at four to six cycles/hour. After the concentrated extract and rinse were transferred extraction was completed, the extract was to a 10-mL volumetric flask for HPLC analysis. cooled and dried by passing it through a drying HPLC analysis column containing about 10 cm of anhydrous Phenanthrene (100 mg) was accurately sodium sulfate. The extractor flask and drying weighed and dissolved in methylene chloride to column were rinsed with a further 125 mL of make a 100-mL internal standard solution of methylene chloride. The dried extract was 1000 mg/L concentration. To the concentrated concentrated to less than 10 mL in a K-D soil extract was added 1.0 mL of internal apparatus in about two hours. No solvent standard solution. The final analysis solution exchange was performed. Once the K-D was made up by adding methylene chloride to apparatus was cooled, the Snyder column was complete a 10-mL volume. A 1 µL aliquot of removed and an additional 2 mL of methylene this solution was analyzed by chiral HPLC using chloride were used to rinse the flask. The

Table 1. Summarized results of the separation of pesticide stereoisomer mixtures using the chiral method development kit.

CHIREX MALA- METOL- NAPRO- FONOFOS PEERMETHRIN PROPARGIT COLUMN THION ACHLOR PAMIDE

2 overlapped peaks αa = 1.432 (30% 2 peaks (α = (α=1.1) and 2 No No No ISP/Hb; 0.5 1.28) (0.25% 3001 separated peaks separation separation separation mL/min) ISP/H, 0.3 (α=1.15) (0.25% mL/min) ISP/H, 0.3 mL/min)

2 overlapped 2 overlapped peaks peaks (α = 1.03) 3005 (α=1.14) and 2 No No No and 1 separated (NP)c No separation separated peaks separation separation separation peak (0.25% (α=1.13) (0.5% ISP/H, 0.5 ISP/H, 0.4 mL/min) mL/min)

3005 No No No Nno separation Nno separatio No separatio (RP)d separation separation separation

cis /trans α=1.25 (15% 2 overlapped peaks 3010 No No No (α=1.11) (5% ISP/H, 0.5 (α=1.34) (1% (NP)c separation separation separation ISP/H, 0.5 mL/min) ISP/H, 0.5 mL/min) mL/min)

α=1.09 (95% 3010 No No No acetonitrile/water, Nno separation No separatio (RP)d separation separation separation 0.4 mL.min)

No No No 3014 Nno separation Nno separatio No separatio separation separation separation

2 sets of overlapped cis /trans No No No peaks (α=1.16, 3020 No separation (α=1.19) (0.5% separation separation separation 1.30) (0.5% ISP/H, 0.4mL/min) ISP/H, 0.4 mL/min) a. α: separation factor b. Solvent used in normal-phase separation is isopropanol (ISP) in hexane (H). For example, 1% ISP/H represents 1% v/v of isopropanol in 99% of hexane. c. NP: Normal-Phase Mode d. RP: Reverse-Phase Mode

40 Proceedings of the 2000 Conference on Hazardous Waste Research Figure 3. The chromatogram of the separation Figure 4. The chromatogram of the separation of Permethrin enantiomers on a Chirex 3005 of Propargite enantiomers on a Chirex 3001 column using 0.25 % isopropanol in hexane. column using 0.3% isopropanol in hexane. a Chirex 3001 analytical column with 30% hexane ring (Welch and Szczerba, 1998). The isopropanol in hexane as the mobile phase. four trans-isomers (Figure 1) were separated on Chirex 3001 using 0.3% isopropanol in RESULTS AND DISCUSSIONS hexane as shown in Figure 4. Only one pair of Six chiral pesticides were screened with Propargite enantiomers was baseline separated. the chiral method development kit to find the Napropamide enantiomers were optimal separation conditions for each. The baseline separated on both Chirex 3001 and results are summarized in Table 1. Isopropanol 3010 columns, using isopropanol and hexane as and hexane were the only two solvents tried in elutes. Figure 5 shows the chromatogram of the normal-phase chromatography, and methanol resolution of Napropamide on a Chirex 3001 and acetonitrile were used in reverse-phase column using 30% isopropanol in hexane. systems. In all cases normal-phase chromato- Because this was the best separation observed graphic separation was more effective than in any system, Napropamide was chosen for reverse-phase. Different proportions of isopro- our soil recovery studies. panol to hexane played an important role in the As shown in Table 1, in our hands, separations, while flow rate had little effect. Fonofos, Metalochlor, and Malathion could not From the reported diastereoisomer (non- be effectively separated with any of the columns mirror image stereoisomers) ratio (3:1) of of our commercial chiral method development Permethrin, the cis-enantiomers (Figure 1) were kit. The observed degradation of Malathion partially separated on the Chirex™ 3005 further complicated the analysis of this sample column using 0.25% isopropanol in hexanes. as shown in Figure 6. The trans-enantiomers were not separated, as For the pesticides that could be partially or shown in Figure 3. baseline resolved, the instrumental detection Though Propargite has three chiral centers, limit (IDL) was determined for the best-re- only four stereoisomers were observed because solved peak as shown in Table 2. All calibra- of the fixed trans stereochemistry of its cyclo-

Proceedings of the 2000 Conference on Hazardous Waste Research 41 Figure 5. The chromatogram of the separation Figure 6. The chromatogram of the separation of Napropamide enantiomers on a Chirex of Malathion enantiomers on a Chirex 3010 3001 (column kit) column using 30% isopro- column using 2% isopropanol in hexane. panol in hexane.

tion curves of the pesticides had R2 (R: correla- the surrogate because of its structural similarity tion coefficient) values greater than 0.98. to the analyte. The recovery of a surrogate can To evaluate the performance of the total also be used to monitor unusual matrix effects analytical process, a clean matrix, e.g. organic- and sample processing problems. free reagent soil, was spiked. Local soil was In general, Method 3540C, Soxhlet collected, washed with water and methanol, and extraction, is considered the standard procedure air dried. Further oven drying only reduced the for analyzing a broad range of solid samples and weight of the soil samples by 0.5%. provides acceptable extraction efficiency for Because there was no suggested concen- most analytes. Soil samples in this study were tration or surrogate standard for Napropamide extracted for 24 hours. After concentrating the HPLC analysis by Method 3500, a 1-mL extracts, HPLC analysis was used to determine sample of 1000 mg/L (ppm) Napropamide the amount of Napropamide recovered. The stock solution and 1 mL of 1000 mg/L naphtha- recovery rates of surrogate and Napropamide lene surrogate standard solution were added to enantiomers were greater than 90%. The this soil sample. Naphthalene was selected as results are summarized in Table 3. Since the Table 2. Instrumental detection limit (IDL) and linear range for three separable pesticide stereoi- somer mixtures using a chiral method development kit.

P)esticides I)DL (ppm Linear Range (ppm

Napropamidea 00.1 0.1-50

Permethrinb 500 50-50

Propargitec 10.0 1.0-50 a. 1st isomer peak Figure 5 b. trans isomer peak Figure 3 c. 3rd isomer peak Figure 4

42 Proceedings of the 2000 Conference on Hazardous Waste Research All four Propargite stereoisomers were sepa- rated. Along with finding the optimal separation conditions for three chiral pesticide stereosiomer mixtures, the instrumental detection limit and linear detection range were measured. The calibration curve within the linear detection range had an R2 greater than 0.98. As expected, the longer ChirexTM 3001 Figure 7. The chromatogram of Napropamide analytical column was more efficient than the on an analytical Chirex 3001 using 30% shorter kit column in separating Napropamide isopropanol in hexane but not to the extent of four times the cost. recovery rate of the surrogate was greater than Nevertheless the chiral method development kit 94%, no unusual matrix effects were observed. was effective in guiding the purchase of the  Two different sizes of Chirex 3001 analytical column for optimum performance. columns were used to separate Napropamide EPA SW-846 Method 3540C was enantiomers: a 50 x 3.2 mm (chiral method shown to be effective in recovering greater than development kit, Figure 5) and a 250 x 4.6 mm 90% of Napropamide from spiked soil samples. (analytical column, Figure 7). The performance Determination of the method detection limit of both columns was compared based on their (MDL) is still needed. α separation ( ) and resolution (Rs) factors. The methods developed in our work may These results are reported in Table 4. be used to study the difference of enantiomeric CONCLUSIONS bioactivities, biodegradation, accumulation, and A chiral method development kit was distribution in soil, surface water, and ground- able to separate three of the six selected chiral water, and to educate others on the importance pesticides. The enantiomers of Napropamide of this problem. were baseline separated. Three peaks of the four Permethrin stereoisomers were observed.

Table 3. Recovery rates of Soxhlet extraction of surrogate and Napropamide enantiomers in soil samples. % Recoverya S1urrogate Standard 94.69 ± 4.4 N9apropamide Enantiomer 1 92.46 ± 6.3 N1apropamide Enantiomer 2 90.91 ± 5.6 a. 95% Confidence Level

Proceedings of the 2000 Conference on Hazardous Waste Research 43 ACKNOWLEDGMENTS Arbor Press, Chelsea, Michigan, p. 3. We acknowledge the financial support Barceló, D., 1991, Occurrence, Handling and of the U.S. EPA (Grant EPA-ERC 2K-R2) Chromatographic Determination of through the Center for Environmental Resource Pesticides in the Aquatic Environment: Management at the University of Texas at El A review, Analyst, 116, pp. 681-689. Paso. Additional funding was provided by the Buser, H., and M.D. Müller, 1995, Environ- Department of Chemistry and the Camille and mental Behavior of Acetamide Pesticide Stereoisomers. 1. Stereo and Henry Dreyfus Foundation (SG-99-078). We Enantioselective Determination Using also gratefully acknowledge the loan of equip- Chiral High-Resolution Gas Chroma- ment by the El Paso Natural Gas Company. tography and Chiral High-Performance Liquid Chromatography, Environ. Sci. REFERENCES Technol., 29, pp. 2023-2030. Ariens, E.J., J.J.S. van Rensen, and W. Welling, Falconer, R.L., T.F. Bidleman, D.J.Gregor, 1988, Stereoselectivity of Pesticides. R. Semkin, and C. Teixeira, 1995, Biological and Chemical Problems, Enantioselective Breakdown of a- Elsevier Science Publishing, New York, hexachlorocyclohexane in a Small NY., Chapter 4-10. Arctic Lake and Its Watershed, Barbash, J. E., and E.A. Resek, 1996, Pesti- Environ. Sci. Technol., 29, pp. cides in Groundwater. Distribution, 1297-1302. Treads, and Governing Factors, Ann

Table 4. Conditions, separation, and resolution factors for Napropamide resolution by short kit and analytical columns.

Snhort Column Analytical Colum

Dmimension 5m0 mm x 3.2 m 250 mm x 4.6 m

Chromatography Settings

Msobile Phase 3s0% isopropanol in hexane 30% isopropanol in hexane

Fnlow Rate 0n.5 mL/mi 1.0 mL/mi

Separation Factor (α)a 15.429 1.43

Resolution Factor (Rs)b 24.00 2.4

a. − α=TTrB M − TTrA M

TrA and TrB are the retention time of the two enantiomers, TM is the retention time of the solvent front. b. 2(TT− ) Rs = rB rA + ()WWAB

WA and WB are the width of the peaks of the two enantiomers.

44 Proceedings of the 2000 Conference on Hazardous Waste Research Faller, J., H. Huhnerfuss, W.A. Konlg, R. Phenoxycarboxylic Acid Herbicides by Krebber, and P. Ludwlg, 1991, Do Marine Microorganisms, Chemo- Marine Bacteria Degrade a- sphere, 24(10), pp. 1423-1429. Hexachlorocyclohexane Milne, G.W.A, 1995, CRC Handbook of Stereoselectively?, Environ. Sci. Pesticides, CRC Press, Ann Boca Technol., 25, pp. 676-678. Raton, Florida. Farran, A., S. Ruix, C. Serra, and M. Aguilar, Müller M.D., and H. Buser, 1995, Environmen- 1996, Comparative Study of High- tal Behavior of Acetamide Pesticide Performance Liquid Chromatography Stereoisomers. 2. Stereo and and Micellar Electrokinetic Capillary Enantioselective Degradation in Sewage Chromatography Applied to the Sludge and Soil, Environ. Sci. Analysis of Different Mixtures of Technol., 29, pp. 2013-2037. Pesticides, J. Chromatography A, 737, pp. 109-116. Renner, R., 1996, Chiral Compounds Show Promise As Environmental Tracers, Grosser, A. A., J.F. Ryan, and M.W. Dong, Environ. Sci. Technol./News, 30, 1993, Environmental Chromatographic No.1, pp. 16A-17A. Methods and Regulations in the United States of America, J. Chromatog., Tedd, V. A., A.J. Willetts, and H.M. Lappin- 642, pp. 75-87. Scott, 1994, Enantioselective Degrada- tion of the Herbicide Mecoprop[2-(2- Iwata, H., S. Tanabe, T. Iida, N. Baba, J.P. methyl-4-chlorophenoxy)propionic Ludwig, and R. Tatsukawa, 1998, acid] by Mixed and Pure Bacterial Enantioselective Accumulation of a- Cultures, FEMS Microbilogy Ecology, hexachlorocyclohexane in Northern 14, pp. 191-200. Fur Seals and Double-Crested Cor- morants: Effects of Biological and Venis, M.A., 1982, The Rational Design of the Ecological Factors in the Higher Tropic Optically Active Wild Oat Herbicide L- Levels, Environ. Sci. Technol., 32, Flamprop-isopropyl, Pestic. Sci., 13, pp. 2244-2249. pp. 309-317. Kurihara, N., J. Miyamoto, and H.A. Kuiper, Welch, C. J., and T. Szczerba, 1998, HPLC 1997, Pesticides Report 37: Chirality in Separation of the Enantiomers of Synthetic Agrochemicals: Bioactivity and Commercila Agricultural Compounds Safety Consideration, Pure and Applied Using the Whelk-O CSP, Enantiomer, Chemistry, 69, pp. 2007-2025. 3, pp. 37-43. Ludwig, P., W. Gunkel, and H. Huhnerfuss, Understanding Environmental Methods, 1992, Chromatographic Separation of Version 2.0, 1998, CD-ROM, Genium the Enantiomers of Marine Pollutants. Publishing Corporation, Amsterdam, Part 5: Enantioselective Degradation of New York.

Proceedings of the 2000 Conference on Hazardous Waste Research 45